measurements of methane emissions from biogas production

MEASUREMENTS OF METHANE EMISSIONS FROM BIOGAS PRODUCTION REPORT 2015:158

ENVIRONMENT

Measurements of methane emissions from biogas production Data collection and comparison of measurement methods MAGNUS ANDREAS HOLMGREN, MARTIN NØRREGAARD HANSEN, TORSTEN REINELT, TANJA WESTERKAMP, LARS JØRGENSEN, CHARLOTTE SCHEUTZ, ANTONIO DELRE

ISBN 978-91-7673-158-1 | © 2015 ENERGIFORSK Energiforsk AB | Telefon: 08-677 25 30 | E-post: [email protected] | www.energiforsk.se

MEASUREMENTS OF METHANE EMISSIONS FROM BIOGAS PRODUCTION

Authors‘ foreword This project was executed mainly by the authors     

Magnus Andreas Holmgren (SP) Martin Nørregaard Hansen (Agrotech) Torsten Reinelt and Tanja Westerkamp (DBFZ) Lars Jørgensen (DGC) Charlotte Scheutz and Antonio Delre (DTU)

and the measurement teams from AgroTech, DBFZ, DGC, DTU and SP. Invaluable contributions were also given by members of the project group        

Karine Adam, Ineris David Baxter, EC JRC & IEA Task 37 Angelika Blom and Caroline Steinwig, Avfall Sverige Bernd Linke, ATB Sören Nilsson-Påledal, Tekniska Verken i Linköping Olga Oliveti Selmi and Zacaria Redad, GDF Suez/Engie Tobias Persson, Energiforsk Zacaria Redad, GDF Suez/Engie

Magnus Andreas Holmgren acted as project leader, and during his parental leave Johan Yngvesson, Gunilla Henriksson and Johanna Björkmalm at SP all made strong contributions to the project. The work started in January 2014, with some main activities being performed in September 2014, namely a workshop on methane emissions arranged by IBBA and comparative measurements at a biogas plant in Linköping, Sweden. This project was funded by the Swedish Energy Agency through Energiforsk – Swedish Energy Research Centre (former SGC). The project was also funded by Cosvegaz SA, GDF Suez/Engie, E.ON Gas Sverige, Stockholm Gas, Tekniska Verken i Linköping, Kraftringen and Avfall Sverige. In-kind contributions were given by DBFZ, DGC, DTU, Ineris and Tekniska Verken i Linköping. We would like to especially thank Tobias Persson at Energiforsk for his strong dedication to the subject and for working so hard in making this project come true.

3


Sammanfattning Detta projekt samlar några av de mest erfarna organisationerna och människorna i Europa när det gäller mätning av metanutsläpp från produktionsanläggningar för biogas. Projektet har två huvudsyften. Dess första syfte är en litteraturstudie för att samla befintlig data och kunskap, främst i Europa, när det gäller metoder för att mäta metanutsläpp och deras resultat. Dess andra syfte är en jämförelse av några mer eller mindre etablerade mätmetoder under en jämförande mätning på en biogasanläggning. I vår kännedom har inte något liknande arbete utförts tidigare. Målet med projektet är att resultaten kan ligga till grund för ytterligare projekt, såsom en gemensam europeisk standardisering av mätmetoder och datautvärdering. Standardisering är viktigt för att säkerställa att anläggningars prestanda bedöms på samma sätt när de installeras i olika länder. Biogas anses vara ett klimatneutralt bränsle eftersom kolet i biogasen binds från atmosfärisk koldioxid, CO2. Den CO2 som frigörs vid förbränning av biogas anses därför vara biogen och inte fossil. Vidare så ersätter biogasen förbrukning av fossila bränslen vilket sänker de totala CO2-utsläppen. Biogas består huvudsakligen av metan (CH4), och eftersom metan i sig är en stark växthusgas så är det viktigt att samla kunskap om metanutsläppen som sker i form av förluster som kan uppstå i produktionskedjan för biogas, och det är också viktigt att minimera dessa utsläpp. Metan är en växthusgas och bidrar till den globala uppvärmningen. Räknat per utsläppt ton så bidrar metan med 28-34 gånger mer påverkan jämfört med CO2. Utsläpp av ett ton metan motsvarar alltså 28-34 ton koldioxidekvivalenter. När man diskuterar metanutsläpp från biogasproduktion är det dock viktigt att komma ihåg att metanutsläpp också förekommer från naturgasanläggningar, vid alternativ behandling av gödsel och från deponier. Diskussioner har pågått i många delar av Europa och världen om det här problemet och flera initiativ för att bestämma storleken på metanförlusterna från biogasproduktion har gjorts. Mätningar har utförts med hjälp av många olika metoder och angreppssätt, och utsläppssiffror baserade på rena antaganden har också påträffats i litteraturen. I biogas- och uppgraderingsanläggningar kan utsläpp av metan till luft förekomma i olika delar av systemet. Vanliga utsläppskällor på biogasanläggningar är ventilationer, buffert/lagertankar, rötkammare, biogödsellager och avvattningssystem. Vanliga utsläppskällor på uppgraderingsanläggningar är koldioxidrik restgas och ventilationsluft med spår av metan, samt anläggningens analysinstrument. Förutom det faktum att metan är en växthusgas så finns det en rad andra skäl till varför dessa utsläpp bör minimeras, inkluderande säkerhetsaspekter, anläggningsekonomi och lukt. Den genomförda litteraturstudien sammanfattar ett antal studier av metanutsläpp på biogasanläggningar i ett antal olika länder. Från litteraturstudien kan man dra slutsatsen att utsläppsstudierna har genomförts med varierande metoder och angreppssätt. Den stora variationen i metoder gör att det är svårt att dra några generella slutsatser utifrån de tillgängliga resultaten. Vad som är en typisk

4

biogasanläggning varierar också mellan olika länder vilket gör jämförelsen ännu svårare. Rapporterade resultat av de totala metanförlusterna från biogasanläggningar sträcker sig typiskt mellan 1 - 3 % av den producerade mängden metan. Det finns i allmänhet två fundamentalt olika angreppssätt för att mäta utsläpp av gaser från ett område eller en anläggning som har diffusa utsläppskällor, t.ex. en biogasanläggning. Dessa olika angreppssätt är fjärranalys och mätningar på plats. Fjärranalys kan utföras med ett antal olika metoder som samtliga är inriktade på att bestämma hela anläggningens utsläpp genom att mäta metankoncentrationen i omgivningsluften vid anläggningen. I detta projekt prövades två olika tekniker för fjärranalys; dispersionsmodellering med bLs-modell (”backwards Lagrangian stochastic”) och spårgasmetoden. Dispersionsmodellering baseras på mätningar av den genomsnittliga metankoncentrationen längs olika mätsträckor i vindriktningen nedströms anläggningen, samtidigt som olika väderförhållanden i omgivningen uppmäts. För detta ändamål kan ett s.k. Open-Path-system (t.ex. FTIR eller laser) användas tillsammans med meteorologiska mätningar. Spårgasmetoden kombinerar ett kontrollerat utsläpp av en spårgas från den studerade anläggningen med tidsupplösta koncentrationsmätningar i vindriktningen nedströms anläggningen. Eftersom denna metod använder en spårgas för att bestämma luftdispersionen så behövs inte någon matematisk modellering av denna. Fördelen med fjärranalys är att den totala mängden utsläpp från anläggningen bestäms, men metoderna kan inte användas för att lokalisera enskilda utsläppskällor eller kvantifiera deras andel av de totala utsläppen. Mätningar på plats utförs genom att var och en av utsläppskällorna bestäms enskilt vid respektive källa. Med detta angreppssätt är det möjligt att skilja mellan stationära (t.ex. ventilation) och diffusa (t.ex. läckage) utsläppskällor. De enskilda utsläppens andel av de totala utsläppen kan också bestämmas. I detta projekt provades både direkta mätningar av metanhalt och gasflöde i utsläppspunkterna, samt provtagningssystem som provtar en stor gasvolym genom utspädning av utsläppet med luft. Speciella provtagningstekniker har också tillämpats vid exempelvis biogödsellager och biofilter, där delar av ytan täcks in med provtagningshuv. Totalt fyra olika team genomförde mätningar på plats i samband med den jämförande mätningen i Linköping. Den jämförande mätningen inleddes också med att tre olika team genomförde läcksökning på anläggningen. Två av teamen hade tillgång till varsin IR-kamera för detektion av metanutsläpp som annars inte är synliga för blotta ögat. Det tredje teamet hade endast tillgång till ett traditionellt läcksökningsinstrument som måste föras intill ett läckage för att det ska kunna detekteras. Slutsatsen av jämförelsemätningarna i Linköping är att de allmänna resultaten från olika metoder och arbetssätt är jämförbara. I Figur 1 ses resultat från de fyra team som mätte på plats på anläggningen. I Tabell 2 och Tabell 3 sammanfattas resultat från mätning på plats och med fjärranalys för två olika tidsperioder. Den studerade anläggningen är stor i storlek och de totala utsläppen är jämförelsevis låga. Den totala metanförlusten som rapporterats från mätningarna på plats i anläggningen varierar mellan 0,6-1,1 % av den producerade mängden metan. Det finns höga och okända osäkerheter i alla mätresultat och de beror på både analytiska osäkerheter och tidsvariation i utsläppskällor. Några lärdomar gjordes från de jämförande mätningarna i Linköping. När schemat för mätningarna sattes inrättades det avsiktligt för att inte ha flera team i samma mätpunkt samtidigt som kunde påverka eller störa varandra. På grund av den stora

tidsvariationen i utsläppskällorna har detta val dock gjort att en strikt jämförelse av resultaten inte kan göras. Variationer i driften under mätveckan inträffade som gjorde det ännu svårare att jämföra resultaten. För en framtida liknande mätkampanj rekommenderas det därför att så många mätningar som möjligt utförs parallellt, och mätningar bör kompletteras med internkalibreringar med gasflaskor med känd metanhalt. Ett föreslaget nästa steg är att ta fram en gemensam handbok om mätningar av metanutsläpp från biogasanläggningar. Handboken ska hjälpa användaren att välja en mätmetod och ett angreppssätt anpassat till syftet med mätningen. Handboken bör lista för- och nackdelar för respektive metod och arbetssätt. Vidare bör den vägleda användaren i att analysera och förstå rapporterade värden från olika metoder och angreppssätt. Handboken skulle också vara en viktig referens vid ett framtida möjligt arbete med standardisering av metanutsläppsmätningar från biogasanläggningar.

1 000 900 Metanemission (g CH4/h)

800

Analysinstrument

700

Kompressorhus 2

600

Kompressorhus 1

500

Aminskrubber Aktivt kol-filter

400

Biofilter 300

Hygieniseringstank

200

Homogeniseringstank

100 0 AgroTech

DBFZ

DGC

SP

Figur 1. Jämförelse av rapporterade värden för utsläppspunkter av låg magnitud. Utsläpp från rötrestlager och tryckvakter ingår inte i sammanställningen.

Tabell 1. Jämförelse av resultat från mätning på plats och med fjärranalys, tisdag 9 sept.

Utsläppskälla

Värde (kg CH4/h)

Summa utsläppskällor av låg magnitud

0.85

Aminskrubber, tryckvakt utblås, vänstra huset

9.2

Rötrestlager

4.4 / 6.5

Summa på plats mätningar

14.4 / 16.5

Fjärranalys spårgasmetoden (med std. avvikelse)

17.9 ± 3.1

Tabell 2. Jämförelse av resultat från mätning på plats och med fjärranalys, torsdag 11 sept.

Utsläppskälla

Value (kg CH4/h)

Summa utsläppskällor av låg magnitud

0.85

Aminskrubber, tryckvakt utblås, mellan husen

0.32

Rötrestlager

4.4 / 6.5

Summa på plats mätningar

5.6 / 7.6

Fjärranalys bLs modell (med std. avvikelse)

4.9 ± 1.5


Summary This project gathers some of the most experienced organizations and people in Europe regarding the measurement of methane emissions from biogas production plants. The project has two main objectives. The first objective is a literature study to gather existing data and knowledge, mainly in Europe, regarding methane emission measurement methods and results. The second objective is a comparison of some more or less established measurement methods during a joint comparative measurement campaign at a Swedish biogas plant, “Tekniska Verken i Linköping”. To our knowledge similar work has not been performed previously. Biogas is regarded as a climate-neutral fuel since the carbon in the biogas is fixed from atmospheric carbon dioxide, CO2. Biogas consists mainly of methane (CH4), and since methane in itself is a strong greenhouse gas, it is important to gather knowledge about the methane emissions in the form of losses that might occur in the biogas production chain, and subsequently it is important to minimize these emissions. When discussing methane emissions from biogas production, it is important to bear in mind that methane emissions do also occur in natural gas installations, in alternative manure treatment and in landfills. From the literature study it can be concluded that a number of studies of the methane emissions from biogas plants have been performed in different countries, using different methods and approaches. The large variation in methods makes it hard to draw general conclusions from the existing data. A rather large variation between typical plants in different countries makes the comparison even harder. Reported results of the total methane losses from biogas plants typically range between 1 – 3 % of the produced methane. There are in general two fundamentally different approaches to measure gas emissions from an area or facility that have diffuse emission sources, such as a biogas plant. They are remote sensing and on-site measurements. Comparative measurements were performed by six different teams, four of them applied on-site measurement methods and two performed remote sensing methods. The conclusion of the comparative measurements is that the general results from different methods and approaches are comparable. The studied plant is large in size and the overall emissions are comparably low. The total methane loss reported by the on-site measurement teams range between 0.6 – 1.1 % of the produced methane. There are high and unknown uncertainties in all measurement results and they are due to both analytical uncertainties and time variation in emission sources during the testing period of one week. A suggested next step would be the production of a handbook on methane emission measurements from biogas installations. This handbook should aid the user in choosing a suitable measurement method and approach, depending on the purpose of the measurement task. It should list advantages and disadvantages of the respective methods and approaches. Further, it should guide the user in analyzing and understanding reported values using different methods and approaches. The handbook would serve as an important reference to future work on standardization of methane

8

emission measurements from biogas installations. Standardization is important in order to ensure that plant performance is deemed likewise when installed in different countries.


Table of contents 1

Background 1.1 Goal of the project

14 14

2

Measurement methods 2.1 Remote sensing

16 17

2.2 2.3

3

17

2.1.2 The tracer dispersion method Leak detection On-site measurements 2.3.1 Methane concentration 2.3.2 Direct measurement of gas flow

18 19 19 19 20

2.3.3 Open area measurements 2.3.4 High volume sampler technique

20 22

Biomethane production and methane emission measurements in the studied countries 3.1 Biomethane production

3.2

4

2.1.1 The inverse dispersion modelling method

24 24

3.1.1 Austria

24

3.1.2 Denmark 3.1.3 France

25 25

3.1.4 Germany 3.1.5 Sweden 3.1.6 Switzerland

26 26 27

Existing measuring programs and performed measurements 3.2.1 Austria 3.2.2 Denmark 3.2.3 France

27 27 28 29

3.2.4 3.2.5 3.2.6 3.2.7

30 35 36 40

Germany Sweden Switzerland Other countries

Measurements at Tekniska verken in Linköping, Sweden 4.1 Plant description 4.2 Leak detection 4.3 On-site emission measurements 4.3.1 Homogenization and hygienization tanks 4.3.2 Biofilter 4.3.3 Digestate storage 4.3.4 Activated carbon filter building 4.3.5 Chemical scrubber, CO2 release

10

41 42 43 49 49 52 53 55 55


4.4

4.5 5 6

4.3.6 Chemical scrubber, compressor buildings

56

4.3.7 Chemical scrubber, pressure relief vents 4.3.8 Analysis instruments on site 4.3.9 Comparison of on-site measurements Remote sensing emission measurements 4.4.1 The backward Lagrangian stochastic method 4.4.2 The tracer dispersion method

57 57 58 60 60 61

4.4.3 Comparison of remote sensing results Comparison of on-site and remote sensing results

61 65

Conclusion

67

5.1

67

Continued work

Literature

68

Annex A. AgroTech Identification and quantification of methane emission from leaks at biogas plants B. DBFZ - on-site method Methane emissions from a biogas production site in Linköping C. DBFZ - remote sensing method Methane emissions from a biogas production site in Linköping D. DGC Methane emissions from biogas plant E. DTU Quantification of fugitive methane emissions from the biogas plant in Linköping F. SP Methane measurements at the Linköping biogas plant

11


Abbreviations Abbreviation

Explanation

ATEX

Appareils destinés à être utilisés en ATmosphères EXplosibles. 2 EU directives to protect employees from explosion risk in areas with an explosive atmosphere.

bLs

backward Lagrangian stochastic (model).

CHP

Combined Heat and Power (unit).

CRDS

Cavity Ring-Down Spectroscopy.

DBFZ

Deutsches Biomasseforschungszentrum. German Biomass Research Centre.

DGC

Dansk Gasteknisk Center. Danish Gas Technology Centre.

DTU

Danmarks Tekniske Universitet. Technical University of Denmark.

FID

Flame Ionization Detector.

FTIR

Fourier Transform InfraRed spectroscopy.

HBT

Hohenheimer Biogas yield Test.

IEA

International Energy Agency.

IR

InfraRed spectroscopy.

ISO

International Organization for Standardization.

LEL

Lower Explosive Limit. The lowest concentration of a gas in air capable of producing a flash of fire in presence of an ignition source. Methane gas has a LEL of 4.4 vol%.

OTNOC

Other Than Normal Operating Condition.

ppm

parts per million

SP

Technical Research Institute of Sweden.

TDLAS

Tunable Diode Laser Absorption Spectroscopy.

VDI

Verein Deutscher Ingenieure. Association of German Engineers.

12


Definitions Phrase

Definition

Active area source

An area source of emission with gas flow being actively driven off, typically an open biofilter where a fan controls the gas flow through the filter.

Closed chamber

Sampling hood which encloses the sampling area and does not allow for air passing in or out of the chamber, also called static chamber.

Open chamber

Sampling hood which encloses the sampling area and allows for air entering the chamber with a known mass flow and allows for air leaving the chamber, also called dynamic chamber.

Passive area source

An area source of emission where gas flow is not being actively driven off.

13


1

Background

Biogas is regarded as a climate-neutral fuel since the carbon in the biogas is fixed from atmospheric carbon dioxide (CO2). The CO2 released when combusting biogas is therefore regarded as being biogenic rather than fossil. Further, any consumption of fossil fuels replaced by biogas will lower the total CO2 emissions. Biogas consists mainly of methane (CH4), and since methane in itself is a strong greenhouse gas, it is important to gather knowledge about the methane emissions in the form of losses that might occur in the biogas production chain, and subsequently it is important to minimize these emissions. Methane is a greenhouse gas and contributes to global warming. Calculated per emitted ton, methane contributes 28-34 times more compared to CO2. Thus, a methane emission of one ton corresponds to 28-34 tons of carbon dioxide equivalents [1]. When discussing methane emissions from biogas production, it is important to bear in mind that methane emissions do also occur in natural gas installations, in alternative manure treatment and in landfills. The Renewable Energy Directive (Directive 2009/28/EC) was adopted in 2009 by the EU. According to the Directive, by 2020, all member states shall have 10 % (on energy basis) renewable fuels in the transport sector. A number of sustainability criteria must be met in order for a biofuel to be accounted. To be accounted as sustainable the greenhouse gas emission savings from the use of the biofuel must be at least 35 % compared with the use of a fossil fuel. Over time the greenhouse gas constraint increase, the savings must be 50 % by the year 2017 and from the year 2018, biofuels produced in plants taken into operation after January 1st 2017, the savings must be at least 60 % [2]. Consequently the sustainability criteria are closely related to methane emissions and measurements of these at biogas plants. Discussions have been ongoing in many parts of Europe and the world about this issue and several initiatives to determine the size of the losses have been made. Measurements have been performed using many different approaches and methods, and methane emission figures based on assumptions have also been found in literature. In biogas and upgrading plants methane emissions to air can occur in different parts of the system. Common sources of emissions at biogas producing plants are ventilation, buffer/storage tank, digester, digestate storages and dewatering equipment. Common sources of emissions at biogas upgrading plants are the off-gas, ventilation and analysis instruments. Except for the fact that methane is considered a greenhouse gas there are other reasons to why these emissions should be minimized and these are: safety aspects, economy and odor [3]. 1.1

GOAL OF THE PROJECT

The project has two main objectives. The first objective is a literature study to gather existing data and knowledge, mainly in Europe, regarding methane emission measurement methods and results. The second objective is a comparison of some more

14


or less established measurement methods during a joint comparative measurement campaign at a Swedish biogas plant, Tekniska verken in Linköping. To our knowledge similar work has not been performed previously. The goal of the project is that the results can serve as a basis for further projects, such as a common European standardization of measurement procedures and data evaluation. The latter is seen as particularly important in order to make available measurement results comparable. Standardization is important to ensure that plant performance is deemed likewise when installed in different countries.

15


2

Measurement methods

In general, there are two fundamentally different approaches to measure gas emissions from an area or facility that have diffuse emission sources, such as a biogas plant. They are remote sensing and on-site measurements. Remote sensing measurement methods include a number of different methods all aiming for quantification of whole site emissions by measuring atmospheric methane concentrations often off-site but not always. In this project two remote sensing methods were tried out; the backward Lagrangian stochastic (bLs) model and the tracer dispersion method. The bLs model is based on measurements of average methane concentrations along different paths downwind the facility together with simultaneously measuring the meteorological conditions in the surroundings. For this purpose an Open-Path-System (e.g. FTIR or laser) can be used along with one or more meteorological stations. With this data the emission rate can be simulated with an inverse dispersion model or radial plume mapping. The tracer dispersion method combines a controlled tracer gas release from the treatment facility with time-resolved concentration measurements downwind of the facility [4]. As the method applies a tracer gas to determine the air dispersion, modelling is not needed. The advantage of both methods is that they determine the total amount of emission from the full scale facility – integrating the individual emissions from different on-site sources/leak areas. Thus the methods cannot be used to locate single on-site emission sources or quantify their share of the total emission. On-site measurement methods aim to identify and quantify single sources of emission at the site. This approach enables the differentiation between stationary (e.g. gas upgrading) and diffuse (e.g. leaks) sources of emission. Some advantages and disadvantages of these methods are listed in Tabell 3. Tabell 3. Advantages and disadvantages of remote sensing and on-site measurement

Advantages

Remote sensing

On-site

Determination of total emission rate Longer time measurement series No influence on plant operation/design No leakage search and encapsulating Time effort independent of plant size No modeling step involved (tracer disp.)

Localization of single sources Quantification of single sources Low detection limit for total emission rate Weather independent

16


Disadvantages

2.1

Remote sensing

On-site

Depending on weather conditions (wind direction and wind speed) Disturbance by topography, buildings, trees, etc.. (especially relevant for bLs) Uncertainties of inverse dispersion modelling (only bLs) Requires site access and prober tracer placement (only tracer dispersion) Requires drivable roads in prober distance to the site (only tracer dispersion)

Only short time window per single source Possibly not all sources accessible Time consuming on large plants

REMOTE SENSING

To date no international standard exists for the measurement of fugitive and diffuse methane emissions from anaerobic digestion facilities. However, there is a basic guideline on the investigation of diffuse sources (VDI 4285) where remote sensing methods are considered. In addition, a mandate from the European Commission to CEN has been issued to standardize measurement techniques for the emission monitoring of diffuse emissions from industrial sites, such as refineries [5]. Several remote sensing methods are descripted in detail by the EPA handbook [6].

2.1.1

The inverse dispersion modelling method

Total methane emissions were quantified using an open path spectrometer sensitive to methane in the downwind plume, a meteorological station with 3D sonic anemometer and a backward Lagrangian stochastic model. This model is based on the MoninObukhov similarity theory [7]. The model approximates the emission source as a homogeneous area. If the measurement path is placed far enough downwind this is a viable simplification [8]. The spectrometer measures the path averaged methane concentration on path lengths of up to 500 m. Alternative to TDLAS devices the methane concentration can be determined by Open Path Fourier Transform Infrared Spectroscopy (FTIR) or mobile point measurement devices. The advantage of the open path devices is that a continuous measurement covering the same path is possible over a long time period. A 3D sonic anemometer is essential to register the turbulence parameters in the atmosphere. These are needed for the simulation of the emission rate. Apart from Windtrax other Lagrangian dispersion models are available, e.g. LASAT (Germany) or NAME (United Kingdom). This method has already been used to determine emissions from farms, landfills and biogas plants. The setup is illustrated in Figure 1. For this project DBFZ applied the bLs method for whole site methane quantification, see Annex 3.

17


Figure 1. The setup of TDLAS for bLs modelling of the emission rate from area sources.

2.1.2

The tracer dispersion method

Total methane emissions can be quantified using a mobile tracer dispersion method that combines a controlled release of tracer gas from the biogas facility with concentration measurements downwind of the facility, by using a mobile highresolution analytical instrument ( [9], [10], [11]). The tracer dispersion method is based on the principle that a tracer gas released at a source area, in this case a biogas facility, disperses into the atmosphere likewise the methane emitted from the same area. Since the ratio of their concentrations remains constant along their atmospheric dispersion, the methane emission rate can be calculated using the following expression when the tracer gas release rate is known: 𝑝𝑙𝑢𝑚𝑒 𝑒𝑛𝑑

𝐸𝐶𝐻4 = 𝑄𝑡𝑟 ∗

∫𝑝𝑙𝑢𝑚𝑒 𝑠𝑡𝑎𝑟𝑡(𝐶𝐶𝐻4 )𝑑𝑥 𝑀𝑊𝐶𝐻4 𝑝𝑙𝑢𝑚𝑒 𝑒𝑛𝑑 ∫𝑝𝑙𝑢𝑚𝑒 𝑠𝑡𝑎𝑟𝑡(𝐶𝑡𝑟 )𝑑𝑥 𝑀𝑊𝑡𝑟

where 𝐸𝐶𝐻4 is the methane emission in mass per time, 𝑄𝑡𝑟 is the tracer release in mass per time, 𝐶𝐶𝐻4 and 𝐶𝑡𝑟 are the measured downwind concentrations in parts per billion (ppb) subtracted of their background concentrations and 𝑀𝑊𝐶𝐻4 and 𝑀𝑊𝑡𝑟 are the molar weights of methane and tracer gas, respectively [9].

Figure 2. The principle of the dynamic plume method for quantifying GHG emissions from area sources.

The tracer gas should have a long atmospheric lifetime. Tracer gases often used include N2O, SF6, and acetylene. Downwind plume concentrations are measured driving along transects with an analytical equipment, which is a fast and has a high sensitivity capable of detecting methane and tracer gas concentrations down to ppb level every

18


second ( [9], [11]). A GPS should be connected to the instrument for logging the measured concentrations to their geographical location. Correct simulation of the methane emitted from the source is very important to obtain precise emission rates. In order to obtain the best possible simulation of the source area, the tracer gas should be released from the part of the plant where the most elevated methane concentration are seen and/or expected. The principle is illustrated in Figure 2. For this project DTU applied the tracer dispersion method for whole site methane quantification, see Annex 5. 2.2

LEAK DETECTION

Leak detection, as the name suggests, is performed to detect (to locate) leakages. The methods used for leak detection can not necessarily be used to quantify the size of the emission, or to calculate the mass flow of e.g. methane. The simplest form of leak detection can be performed without technical equipment, using only the human senses. Leaks can be detected by smell, sound, vision or touching a leaking object. With the additional help of leak detection spray, that helps to produce foam or bubbles at the leak, many leaks can be detected and hence hopefully avoided. For performance of systematic leak detection it is however recommended to use a portable leak detection instrument, using a sensor that reacts to a rise in methane concentration. Common simple and cheap leak detection instruments have detection limits of 1-5 ppm, which will enable the user to find very small leaks. Some leak detection instruments will give a value of the methane concentration on a display. Even though the uncertainty of this value is comparably large it is allowed to use in the Swedish voluntary agreement system if the emission in a certain measuring point is less than 0.1 % of the total amount of methane produced in the plant and less than 10 % of the total emissions. Common sensors are semiconductor sensors or catalytic sensors [3]. In recent years methane lasers have been developed. These instruments measure the total methane concentration along the trajectory of a laser beam that is emitted from the instrument, giving the results in a path-integrated unit (ppm m). This technique makes it easier to perform leak detection over large areas. The most advanced leak detection instrument available is the IR camera, which continuously shows an image where leaking methane can clearly be seen in the picture, which is otherwise not visible by the human eye. The IR camera is capable of producing both images and videos. With the aid of an IR camera, large areas can quickly be scanned for leaks. The IR camera is not capable of detecting very small leaks, an approximate limit of detection is 0.5 vol-% methane concentration, but the size of the leak (the mass flow) will also have an influence on the detection limit. 2.3

ON-SITE MEASUREMENTS

2.3.1

Methane concentration

The measurement of methane concentration in waste gases from stationary sources is standardized in EN ISO 25140:2010 and EN ISO 25139:2011. EN ISO 25140:2010 describes the continuous method of using a Flame Ionization Detector (FID), together

19


with a Non-Methane Hydrocarbon cutter which filters out other hydrocarbons than methane. This instrument usually has a wide range of concentration measuring ranges, from 1 ppm up to 100 000 ppm (10 vol-%). This technique has been applied by DGC and SP in this project, see Annexes 4 and 6. The manual sampling method where gas samples are taken in bags for later analysis with a gas chromatograph (GC) in a laboratory is described in EN ISO 25139:2011. This technique has been applied by DBFZ in this project, see Annex 2 (instead of gas bags evacuated glass vials with 22 ml volume were used). Apart from these standardized methods there are also other techniques available to determine the methane concentration such as photo-acoustic, IR and FTIR analysers. The photo-acoustic technique has been applied by AgroTech in this project, see Annex 1.

2.3.2

Direct measurement of gas flow

Determining the gas flow is a great challenge in these measurements since the situation varies a lot between different sampling points and between plants. In addition, the sampling points/sources of emission are rarely prepared for measurements. The main goal is to measure the flow but in many cases it is technically or practically difficult to perform. There are a number of measurement methods to use, for example pitot tube measurements with differential pressure readings, hot wire anemometer, vane anemometer etc.., and a standardized approach is described in EN ISO 16911-1:2013. The technique of direct measurement of the gas flow has been regularly applied by SP in this project, see Annex 6, but other teams also used the technique for some emission sources. If measurements are difficult or impossible to perform there are other ways to determine the gas flow, e.g. the manufacturer’s fan data for a forced ventilation release and as a last resort template values can be used.

2.3.3

Open area measurements

Concentration and flow measurements as described above cannot be performed on open area sources, such as digestate storage tanks or biofilters. Different types of sampling hoods do exist, which can be used to encapsulate a small area of the emission source, enabling measurement and calculation of the total area source emission. The simplest case of such a sampling hood is the sampling hood for biofilters, where the hood has a surface area of typically 1 m2 at the bottom placed on the biofilter surface, and a sampling pipe area of typically ~0,0001 m 2 at the top, see Figure 3. In this way the very low gas flow through the biofilter surface is possible to measure. This type of sampling hood, for active area sources, is standardized by VDI [12]. This type of sampling hood was however not used in this project.

20


Figure 3. Sampling hood for biofilters.

For other area sources, such as digestate storage tanks, sampling hoods based on open chamber and closed chamber techniques can be used (see definitions). For a closed chamber a certain sample area is enclosed by the closed chamber and the gas concentration within the chamber is sampled periodically [13]. The measuring principle is based on the measurement of an increasing gas concentration inside the chamber volume. After putting the closed chamber on the emission surface a gradually increasing gas concentration can be determined. For this project DBFZ applied the closed chamber technique at some emission sources, see Annex 2. For the closed chamber, samples are taken at certain time intervals (e. g. 5, 10, 15, 20, 25, 30 min). Then the emission rate can be calculated from the slope of the gas concentration, the chamber volume and the encapsulated surface area according to Equation 1.

ESpec =

∂c V ∙ ∙ 0.06 ∂t A

Equation 1 ESpec

Surface specific emission mass flow

g m-2 h-1

∂c ∂t

Linear slope of gas concentration

V

Volume inside the chamber

m3

A

Encapsulated surface area

m2

mg m-3 min-1

For an open chamber measurement a known flow of air is passed through the chamber allowing for calculation of the mass flow emission, see Figure 4. Both the open chamber and closed chamber sampling hoods, for passive area sources, are standardized by VDI [14]. For this project, SP applied the open chamber technique at the digestate storage tank, see Annex 6. The open chamber technique can also be applied when measuring the emission from leakages, as has been applied by DBFZ in this project, see Annex 2.

21


Figure 4. Schematic principle of an open chamber.

2.3.4

High volume sampler technique

The basic technique of the high volume sampler is evacuating the leaking biogas from a hole or crevice together with a large amount of diluting air. The sample of diluted biogas is conveyed to a flow measuring device and FID analyzer and the total mass flow of leaking methane is calculated as the product of volume flow and methane concentration above background value in the diluting air. To control the leaking biogas flow the actual leakage (the hole or crevice) is covered by a suction hood or wrapped in plastic sheet, allowing for dilution air and leaking biogas to be conveyed to the flow and concentration measurement. For this project AgroTech and DGC applied the high volume sampler technique at most emission sources, see Annexes 1 and 4.

Figure 5. High volume sampler.

Figure 5 shows the principle of the sampling technique and a picture of a high volume sampler where the sampling hood is attached to a silo tank.

22


The sample system consists of: -

Sampling hood or

-

Antistatic ventilation hose

-

ATEX approved blower

-

Control box including safety circuit in case of too high concentration of methane in the sample gas (alarm breaks in at 10 000 ppm methane equivalent to 25 % of LEL)

-

Flow measurement (calibrated orifice)

-

FID equipped with a Non-Methane Hydrocarbon cutter according to section 2.3.1.

The high volume sampler technique is described elsewhere and may not be exactly the same layout as described above, but the main principle is still the determination of coherent values of sample gas flow and methane concentration. Also portable, intrinsically safe, battery-powered high flow samplers are commercially available. These instruments are mainly designed to determine the rate of gas leakage around various pipefittings, valve seals and compressor seals found in natural gas transmission, storage, and compressor facilities.

23


3

Biomethane production and methane emission measurements in the studied countries

This chapter details the biomethane (biogas) production in studied European countries and also gives an overview of the known methane emission measurements that have been performed. 3.1

BIOMETHANE PRODUCTION

This chapter is based primarily on the publication IEA Bioenergy Task 37 – Country Reports Summary 2014 [15].Unless stated otherwise, this is the source for all information given herein. The biogas market has been established in a number of European countries with significant growth in the 2000 – 2020 period. The largest producers of biogas today are Germany, Italy, UK and France [16]. The largest European producers of biomethane (biogas upgraded to natural gas quality) are Germany, Sweden, Austria, Switzerland and the Netherlands [16]. Countries such as Denmark and UK are growing fast and are currently constructing several biogas upgrading plants due to the introduction of attractive financial incentives. Biogas is produced from biogas in landfills and in anaerobic digesters. The feedstocks used for anaerobic digestion varies in different countries. The main production is based on energy crops in Germany while it is mainly based on various waste products in Sweden and UK [15]. Also the utilisation of the biogas varies between the countries. Electricity production is dominating in Germany and UK while automotive fuel is dominating in Sweden. The utilisation of biogas as automotive fuel is gaining more interest in several countries and is expected to grow significantly in countries such as Denmark, Germany and UK [15]. Below follows a description of the current situation and expected development in the countries that are further discussed in this report.

3.1.1

Austria

Today the main production of biogas is derived from energy crops, sewage sludge and landfills. The annual biogas production corresponds to 1.5–2.5 TWh. Current trends are that high prices of biogas feedstock (e.g. maize) lead to severe difficulties to operate the plants economically. This has led to keen interest to investigate the possibility to use alternative substrates. There is a total of 368 biogas plants in Austria. In Austria biogas is utilised mainly for electricity and heat production. Even though the aim is to upgrade more biogas to biomethane for use as automotive fuel, this change is taking place rather slowly. There are around 7,700 natural gas vehicles (NGVs) and about 180 compressed natural gas (CNG) filling stations. Three of the filling stations are situated at biogas upgrading plants. There are 11 biogas upgrading units in operation. All commercial technologies are represented (amine scrubber, water scrubber,

24


membrane and PSA). The upgrading plants are rather small, 600-800 Nm3/h, and have a combined capacity of around six million Nm3 biomethane annually.

3.1.2

Denmark

154 biogas plants are in operation in Denmark, with a yearly production of 1.2 TWh of biogas. Animal manure is the most important biogas feedstock, with a high prospective potential. According to the Danish Biogas Association, roughly 7% of the animal manure is today supplied to biogas plants in Denmark. The aim is to increase it to 50% by 2020. Along with manure, organic wastes from industry and sewage sludge also make a significant contribution to the biogas production today. In 2012 the Danish Energy Agency predicted a 4-fold increase (to 4.7 TWh) of the total biogas production by 2020. The Biogas Task Force concluded in 2013 that the increase will only be a doubling, to around 3 TWh by 2020. A limited number of biogas projects, representing an increase of about 400 GWh, have already reached a final decision. Today biogas is mainly used for heat and power production in Denmark. The Danish government believes that biogas will be an important automotive fuel in the future, especially when replacing fossil fuels used by heavy duty vehicles. The first four Danish biogas upgrading plants are in operation and a number of biogas upgrading projects are at various stages of planning. There are seven biogas filling stations and more are about to be established. Currently, around 80 CNG cars are in operation in Denmark. In 2013 there was only one small biogas upgrading plant in operation in Denmark that was supplying biomethane to the natural gas network. In 2014, four new biogas upgrading plants were taken into operation. The use of biogas in Denmark as a transport fuel is only in its early days, and the main driving force is government policies in the form of certificates. At the end of 2014 there were 10 filling stations for compressed natural gas (CNG).

3.1.3

France

The vision of the French Environment and Energy Management Agency is to produce 70TWh biogas annually by 2030 and that 600 biogas plants will be built every year. 50% of the biogas produced shall be injected into the grid, 30% shall be used to generate electricity and the remaining 20% shall be used to produce heat. In 2050, the aim is to produce 100TWh. In France there are 256 biogas plants and 245 landfills. The number of farm AD plants is expected to double by the end of 2013. A recent study financed by ADEME on The estimation of feedstock for AD use shows that the potential resources for AD will give a probable potential of 56TWh by 2030. Based on its own calculation, an estimation of ADEME expects a theoretical production of 70TWh by 2030. In France there is a strong development of on-farm and centralized biogas plants and for landfills to recover biogas for electricity generation (today only 90 out of 245 landfills utilize biogas). Around 120 on-farms AD plants were built by the end of 2013 and nearly 15 centralized units. In addition, 60 WWT and 80 agrofood industries AD plants are currently in operation. In 2010, a study showed a relatively low energy recovery from biogas, around 60% of raw energy, the main part coming from landfills. There are only four biogas upgrading plants in operation, but more than 400 applications to inject biomethane into the natural gas grid, which indicate a significant increase of the number of upgrading plants in a nearby future. Today, all the

25


biomethane produced is injected into the natural grid or sold as fuel for vehicles. More than 13,500 vehicles, of which 3,500 are trucks, are in use in France with a daily consumption of 265,000 Nm3. 37 public filling stations and around 130 private filling stations are in operation.

3.1.4

Germany

The share of renewable energies in energy generation is to be raised to 40-45 percent by 2025, to 55-60 percent by 2035, and to 80 percent by 2050. The reform of the Renewable Energy Sources Act (EEG) should play a key role in the success of the energy reforms. The introduction of specific growth targets for different technologies is a new development for the German renewables support scheme. The annual growth of biomass including biogas is limited to a maximum of 100 MW compared to 2,500 MW for onshore wind and solar power. The 7960 biogas plants in the agriculture sector make the biggest contribution to biogas production today with electricity and heat supply of 25,120 GWh/year and 10,550 GWh/year, respectively. According to information from the Federal Ministry for the Environment, Nature Conservation, Building and Nuclear Safety, in 2013 the main part of the biogas was used for electricity and heat production, while biogas utilization as an automotive fuel is rare. The share of energy consumption in Germany for electricity, heat and fuel amounted to 4.7%, 1% and 0.1%, respectively. In 2014 a total number of 151 biogas upgrading plants were in operation with a feed-in capacity to the gas grid of 93,650 Nm3/h biomethane. Compared with data from the previous year, the number of plants increased by 25%. Amine scrubbing, water scrubbing and pressure swing adsorption (PSA) are the most commonly applied technologies. Due to the reduction in the feed-in tariffs, the number of new biogas upgrading plants will presumably be smaller than the 20-30 plants/year erected in previous years. Surveys have revealed that at least five of the projects under construction or in planning are not realized due to the 2014 EEG reform and/or a continuation of the projects are not yet finally clarified. Based on data from Erdgas Mobil GmbH in 2013, about 170 filling stations with 100% biomethane have sold 300 TWh biomethane. This corresponds to 20% of natural gas consumption by the 95,000 registered gas vehicles in Germany.

3.1.5

Sweden

In Sweden there is a governmental aim to produce 50 percent of the energy from renewables by 2020 (this has already been reached), but there are no specific targets for biogas production. Sweden also has a governmental vision to have a fossil free transportation sector by 2050. The results from a public inquiry on how to reach this vision are expected to be important for the future governmental support for biomethane production in Sweden. In Sweden the production of biogas has been fairly constant at around 1.3–1.7 TWh for several years. The main reason is the difficulties in showing a reasonable profit for new investments and new biogas plants. Biogas produced in new plants has been balanced by the steady decline in landfill gas production. In Sweden, around 50% of the biogas is used as CNG/biomethane. This part is increasing every year to meet the increasing demand from the increasing number of gas

26


vehicles. The main part of the remaining biogas is used for heat production. Nearly all upgraded biogas is used as automotive fuel, the annual biomethane production in Sweden is around 900 GWh. The biomethane is produced in 53 biogas upgrading plants with various technologies (~70% water scrubbers and, ~15% PSA, ~15% amine scrubbers). In one plant, with a capacity of 60 GWh, the biomethane is liquefied and sold as LBG (Liquefied Biogas). Of the methane used as an automotive fuel, the biomethane share was 58% on energy basis in 2013. It is used by 47,000 gas vehicles, of which 2,200 are buses and 750 are heavy duty vehicles. Around 210 filling stations dispense CNG/biomethane, five of these also have liquefied gas, LNG/LBG.

3.1.6

Switzerland

The Federal Council has adopted the energy strategy 2050 in order to guarantee security of energy supply in the long term. The thrust of this strategy is to gradually phase out nuclear power and, on the other hand, to develop hydro power as well as the new renewable energies (sun, biomass, biogas, wind, wastes and geothermal heat) and to improve energy efficiency of buildings, appliances and transportation. Energy supply difficulties could be overcome by fossil-fuelled power generation and by energy imports. Concerning biogas, the energy strategy aims an annual electricity production of 1.6 TWh by 2050. In order to reach this goal, the focus is on coordinated energy research. In Switzerland there are around 610 biogas plants and six landfills. The total gross biogas production was 1,129 GWh in 2013. The biogas is mainly used to produce electricity and heat in CHP plants, but the biomethane production is growing rapidly. There are 19 upgrading plants (mainly PSA units, amine scrubbers and organic physical scrubbers), two at agricultural sites, eight at wastewater treatment plants and seven at biowaste AD sites, with a total biomethane production of approximately 128 GWh. The target is to inject 300 GWh into the grid by 2016. Today more than 11,000 vehicles run on methane and 140 filling stations are in operation. 3.2

EXISTING MEASURING PROGRAMS AND PERFORMED MEASUREMENTS

This chapter briefly describes existing measuring programs and other performed measurements in different countries which are known to the project group.

3.2.1

Austria

In Austria a large project called “Klimoneff” was performed in the last years [17]. The focus lay on the quantification of fugitive methane emissions from agricultural biogas plants using a remote sensing method and on the analysis of digestate. The remote sensing method in use was Open Path TDLAS in the downwind plume in combination with inverse dispersion modelling (refer to section 2.1). The emission quantifications were performed on five facilities. One of those was analysed in detail. For this biogas plant the results from seven days of data collecting gave a median value of 4 % loss of the produced methane when the open digestate storage tanks were filled. During periods when the storage tanks were empty, data from 6 measurement days led to quantification of 3 % methane loss. The four other plants were surveyed for two times five hours each. The estimated methane loss during these periods was given within the range from 1.6 up to 5.5 %. At least two of these plants had open digestate

27


storage tanks. For parts of these measurements operational problems of the facilities were reported.

3.2.2

Denmark

Until 2012/13 there have been no studies in Denmark measuring and documenting methane leakages and emissions from biogas plants. In April 2013 AgroTech and DGC initiated a project with the overall aim of reducing greenhouse gas emission from biogas plants in terms of methane losses. This is done by developing a method for identification and quantification of methane leakages. In the project, “Methane Emission from Danish Biogas Plants” funded by Energinet.dk under the ForskEL programme, nine biogas plants were screened for methane emissions and the identified leakages were subsequently quantified. One of the tasks in the Danish project is to test and evaluate the selected method against another method. Participating in the Linköping measurements in this current European project fulfil this demand. The project has developed an operational method for identification and measurement of methane emission from biogas plants. The developed method was utilized to identify and quantify methane emission at ten Danish biogas plants. In total 50 leakages were identified. The measured methane emission from individual leakages varied considerably from zero to 276,000 Nm3 CH4 per year. The total amount of methane emitted from the studied biogas plants summed up to 4.2 % of the total methane production at the studied plants. In 2012/13 DTU Environment performed emission measurements at Avedøre wastewater treatment plant located in the Copenhagen area, Denmark [11]. In total, 9 measurement campaigns were conducted where the total emission of methane and N 2O from the plant was quantified using the tracer dispersion method. The measurements performed on the plant showed that most of the methane emission occurred from the anaerobic sludge treatment, while the majority of the N2O emission derived from the biological nitrogen removal in the aeration tanks. The methane emission rate varied between 5 kg h-1 up to 92 kg h-1, whereas the emission of N2O ranged from the detection limit (0.4 kg h-1) up to 10.5 kg h-1. The high emissions were observed during periods when the plant experienced operational problems. For example, the highest emission of methane was measured during a period with foaming problems in the anaerobic digesters, while the highest N2O emission was observed during sub-optimal operation of biological nitrogen removal in the secondary treatment of wastewater. The measured methane emissions accounted for between 2 and 33% of the total methane production at the plant. Similarly it was seen that between 0.15 (detection limit) and 4% of the total nitrogen into plant was released via N2O emissions. From 2013 to 2015, DTU Environment has measured methane and N2O emissions using the tracer dispersion method from seven wastewater treatment plants, all with anaerobic digesters. The investigations are still ongoing and the results are not yet reported. Measured methane emissions varied between 3 and 30 kg h -1 [18]. In 2013/14, DTU Environment measured methane emissions from three plants treating organic waste (household waste, garden waste and industrial waste) and producing biogas. Two of the plants have combined aerobic / anaerobic treatment, while one plant only performs anaerobic treatment. All three plants showed significant emissions of methane (20 to 40 kg h-1) [18].

28


3.2.3

France

French regulation defines some requirements on methane monitoring: methane concentration in biogas and emissions from combustion plants (engine, turbines, boilers) should be monitored (CH4, CO2, NOx, SOx).

29


Studies have been conducted on anaerobic digestion in France for 2 years in order to identify: -

plant specific feedback on: o

incidents / dysfunctions or OTNOC situations that could lead to methane emissions, and

o

measures that were taken to limit emissions, and to prevent future emissions (continuous improvement).

o

specific hotspot methane emissions, installations or activities where methane could be emitted directly to atmosphere.

The information was collected through a bibliographic review and situations applicable to French plants. In that way, a typology of anaerobic digestion plants was drawn up (for environmental and accidental risk analysis). Several programs to quantify emissions (including methane) are in progress or have started during 2015. The objective is to identify and quantify biogas leakages at several stages [19]. Performed measurements: Stack emissions Measurement campaigns were done in three farm installations on emissions from engines. Results showed that methane emission from engines are in the range of 1,7 – 3,2 % of methane production [20]. Diffuse emissions Some operators have voluntarily done methane emission detection by IR cameras to identify diffuse and fugitive sources. These measurements do not provide quantification but help operators to identify spot emissions and measures to be taken to limit emissions or identify equipment that needs to be monitored or repaired.

3.2.4

Germany

Due to the "Renewable Energies Act" the number of biogas plants in Germany has grown to about 8,000 (agricultural biogas plants with CHP or biogas upgrading and bio-waste treatment plants) [21]. Consequently the avoidance of GHG emissions from biogas plants has been of paramount importance for a few years. Concerning the investigation of GHG emissions, please note that the biogas technology employed at agricultural biogas plants and bio-waste treatment plants differs significantly. Agricultural biogas plants: IR cameras and hand held methane lasers are used allowing remote sensing of leakages from plant components which are difficult to access. There are results showing the occurrence of leakages in gas-conducting plant components on biogas plants. One project investigated ten biogas plants and showed that biogas losses from leakages are relevant [22]. Eight plants had an overall number of 22 leakages and seven of them were evaluated as serious leakages (four occurrences of untight wires to adjust

30


agitators; bull eye (observation window), pipe penetration, leakage in double membrane roof). Another project investigated a large number of biogas plants and evaluated the data concerning the frequency of occurrence of certain leakages [23]. Results from this project are shown in Figure 6. 90

78

Number of leakages

80 70

61

60 50 34

40 30 20

10

5

8

9

13

13

13

17

20

22

25

35

28

0

n = 292

Figure 6. Number and location of detected leakages (modified from [23]).

Typical leakages are the foil roofs (single and double membrane) and the wires to adjust the agitators, but also different leakages in gas-conducting plant components (e. g. CHPs, compressors, gas pipes) can occur. In yet another project, ten agricultural biogas plants in Germany were investigated [24]. From these ten plants seven were based on wet fermentation, two on dry continuous fermentation (no. 1 and 6) and one on dry batch fermentation (no. 5). The measurement program included two measurement periods (summer and winter) in each plant, and every plant component was investigated (substrate storage, feeding unit, digestion, digestate storage and gas utilization). The methane emissions were determined by an on-site method. First a plant survey was carried out to identify the emission spots and in the second step the emissions were quantified. The silage storage, the feeding units (e. g. screw conveyor, mixing tank) and the digestion (e. g. leakages, methane diffusion in the supporting air of double membrane roofs) were investigated by means of an open chamber system. The methane emissions from the open digestate storages were determined by a closed chamber. Not gastight covered storages and substrate storage tanks were investigated by means of the air injection method. The emissions from the gas utilization were directly measured in the exhaust pipe. The measured methane emission factors are shown in Figure 7.

31


Methane emissions in g CH4 kWhel-1

25

20

15

10

5

0 1 Gas utilization

2

3

4

5 6 7 8 9 10 Plant-no. Miscellanaeous Open/not gastight covered digestate storage tank

Figure 7. Methane emissions in g CH4 kWhel-1 from ten agricultural biogas plants in Germany. The emission sources are classified as gas utilization (CHP/upgrading unit), open/not gastight covered digestate storages and miscellaneous [24].

The main emission sources were open or not gastight digestate storage tanks and the gas utilization. While plants 1, 2 and 6 had gastight covered digestate storages and consequently did not show any detectable methane emission from this source, the remaining plants emitted methane amounting to 0.40 – 20.14 g CH4 kWhel-1 (0.22 – 11.22 % CH4 loss) from open/not gastight covered digestate storages. However, the methane emissions from the open/not gastight covered digestate storages were based on two single measurements only. So the results do not represent an average emission factor over a whole year. Furthermore each plant showed a detectable methane slip in the exhaust of the gas utilization. The emission factors from the CHPs varied from 1.09 to 5.89 g CH4 kWhel-1 (0.61 – 3.28 % CH4 loss) for the single CHP units. These methane emissions were caused by an incomplete combustion in the engine. The methane slip from the upgrading units (no. 8 and 9) amount to 9.58 and 2.19 g CH4 kWhel-1 (5.34 and 1.23 % CH4 loss) which were caused by a defective or missing post combustion. Miscellaneous emissions sources (e. g. leakages or the processing of fermentation residues) contributed marginally to the overall methane emissions from the investigated plants. An exception is a leakage from plant 3 which was not considered in Figure 7. This leakage (a not properly closed service opening) caused an emission factor of 5 % CH4-loss. In the last project covered in this section, three agricultural biogas plants with an upgrading step to biomethane were investigated [25]. Each plant was based on wet fermentation and used a chemical scrubber for the upgrading process. Plants 1 and 2 used energy crops and plant 3 used residual materials (distiller's wash). While the digestate storages of plant 1 and 3 were gastight, the storage tanks from plant 2 were not gastight covered. The methane emissions were determined by both an on-site method (comparable procedure to [24]) and a remote sensing method by means of an open-path TDLAS

32


system and the software tool Windtrax. The measurements were carried out simultaneously to ensure the comparability of the results, which are shown in Table 4.

Table 4. Methane emissions in % CH4-loss from three biogas plants with an upgrading step to biomethane in Germany [25]

On-side method

Remote sensing method

MP-Result in % CH4

Uncertainty in % CH4

Average in % CH4

Uncertainty in % CH4

BMA I – MP II

0.12

0.02

0.51

0.21

BMA I – MP III

0.13

2.0

0.8

0.22

0.12

0.02

BMA II – MP I

0.12

1

BMA II – MP II

0.96

1

1.29

BMA II – MP III

0.22

1

-

0.72

1

BMA II – MP V

0.09

1

BMA III – MP I

0.14

0.04

BMA III – MP II

0.13

0.03

BMA III – MP III

0.05

0.01

BMA III – MP IV

0.05

0.01

BMA II – MP IV

0.67

2

0.03

1

0.30

2

2

0.26

1

0.55

2

0.05

1

-

0.22

1

0.66

0.02

1

-

1.33 -

2

2

1

… Summation of only actually measured emission sources … Extrapolation of measured emission factors on the overall system BMA = biomethane plant; MP = Measurement Period 2

From the comparison of the two methods the following conclusions were drawn [25]: 

“Both methods were suitable for quantification of methane emissions of biogas plants.”



The on-site method is time consuming on large sites, but identifies individual sources. Only short periods can be measured at each emission spot. Recurrent measurements during the year may decouple the results from environmental conditions.



“The remote sensing method provides meaningful values over the total emission rate of the plants and their temporal variability. The method is highly dependent on wind conditions and the topological and infrastructural conditions.” Furthermore the measurement does not affect the plant operation.



“At all three plants, the emission rate measured by the remote sensing method was higher than from the on-site method. A comparison of emission rates of two plants, one of them determined by the on-site method and the other by remote sensing, would not be justified due to the systematically lower estimate by the on-site method.”



“Malfunctions can have a significant impact on the total emission level of a biogas or biomethane plant. The continuous measurement of emissions from overpressure/underpressure valves is a promising way to represent the

33


relationship between the operating conditions of biogas plants and the resulting emissions.” Bio-waste treatment plants: In one project twelve German bio-waste treatment plants were investigated [26]. From these twelve plants four were based on wet fermentation, five on dry continuous fermentation and three on dry batch fermentation. The plants were chosen as a representative selection to the German plant inventory. The publication presents the results of the GHG emission measurements and the resulting carbon footprints from the plants. It must be pointed out that bio-waste treatment plants are designed differently compared to agricultural biogas plants. Many plant components (e. g. substrate delivery and pretreatment, processing of fermentation residues, digestate storages, composting tunnel, batch digesters) can be encapsulated and exhausted. Then the exhaust is conducted to a biofilter and is treated there. Consequently each emission source which was conducted to the biofilter is part of this overall emission source. So the single emission sources (e.g. post composting process) cannot be compared between the single plants. The plants are partly very different concerning their process technology and this has to be considered if the methane emissions should be interpreted in the right way. The results show that the emissions are dominated by the manner of plant operation. Figure 8 shows the component-specific emissions (as carbon dioxide equivalents) from each plant. Except for plant 6 (high ammonia emissions caused by non-operating acid scrubber), methane is the most important GHG. Insufficient aeration or unaerated post-composting processes led to increased GHG emissions (especially methane, cp. Figure 8 plants no. 1, 2 and 12). In some plants the GHG emissions from the post-composting are included in “Emissions after biofilter” (cp. Figure 8, e.g. biogas plant 10), because the post-composting process was encapsulated and attached to the biofilter. Furthermore the results show that the open storage of fermentation residues can cause considerable GHG emissions and should be avoided.

34


GHG emissions in kg CO2-eq Mg-1 bio-waste

400 Application of digestate

350 300

Digestate storage tank (liquid/solid residues)

250

Open post-composting

200 150

Emissions after biofilter 100 Electricity demand (biogas plant + CHP)

50

0 1

2

3

4

Wet fermentation

5

6

7

8

Dry fermentation

9

10

11

12

Batch fermentation

CHP emissions (average value for all CHP)

Plant no.

Figure 8. GHG emissions in kg CO2-eq Mg-1 bio-waste from twelve bio-waste treatment plants in Germany (modified from [26]).

Besides diffuse and stationary point sources and area sources, operational and timevariant methane emission sources have to be considered as well. In one study pressure relief vents from two agricultural biogas plants were investigated over long time periods (60 – 100 days) by means of a vane anemometer and temperature sensor permanently installed in the exhaust duct of the pressure relief vent. The methane losses measured were 0.06 CH4-loss (plant A, winter) and 3.88 % CH4-loss (plant B, summer). There were different influences which affect the methane emissions from pressure relief vents; the plant management during normal operation, OTNOC situations (mainly the outage of the primary gas utilization) and the atmospheric conditions. Additionally these influences are able to affect each other and create a compounded impact on the operational methane emissions from pressure relief vents [27].

3.2.5

Sweden

A pair of Swedish studies in 2003 and 2005 showed considerable methane emissions from biogas production plants and biogas upgrading plants. Therefore, in 2007, the Swedish Waste Management Association (Avfall Sverige) introduced a system called the “Voluntary Agreement” for biogas production plants and biogas upgrading plants with the aim to quantify and reduce methane emissions [3]. The purpose of the Voluntary Agreement is to control and minimize the emission of greenhouse gases. Carbon dioxide, methane and nitrous oxide are greenhouse gases that can occur in the biogas system. However, in the Voluntary Agreement system, only methane is included. Methane is the main component in biogas and by focusing on only methane, other factors can be included, such as economy and the safety aspect. The Voluntary Agreement consists of two main parts: 

Systematic leak detection including rectification of found leakages. This is mainly performed by the staff at each plant at least once a year.

35




Emission measurements at discharge points to quantify the emissions and losses. This is performed by an external independent measurement consultant at least once every three years.

The system boundaries of the Voluntary Agreement specify the outer limits for the emission and only emissions within these boundaries are quantified. The following criteria are included [3]: 

Only the parts of the plant that the plant owner owns and can control are included.



Only the parts that are associated with the production of biogas or cleaning/upgrading of the gas are included. Emissions that occur in association with the use of the gas or the digestate, and emissions that occur during transportation of substrate, digestate and gas, are not included in the system.

During the first three-year period of the Voluntary Agreement, 2007-2009, measurements and calculations were performed at 18 biogas producing plants and 20 upgrading plants. For the second period, 2010-2012, 21 biogas producing plants and 28 upgrading plants participated [28]. The average methane emission for the biogas producing plants during the first period was 1.6 % relative to the produced amount of raw gas. A moving average for the time period 2007-2012 was concluded to be 1.9 %. The emission levels seem to be quite constant, and it is more likely that the small increase in methane loss is due to the fact that additional sources of emissions were discovered at the plants during the second measuring period, 2010-2012. In addition, some of the plants that joined for the second period showed relatively high losses of methane [28]. The average methane emission for the biogas upgrading plants was 2.7 % relative to the produced amount of clean gas, for the first period, 2007-2009. The moving average for the time period 2007-2012 was 1.4 % which indicates that substantial improvements had been made at several of the participating plants. Also, newer plants that had joined the system for the second period have modern techniques with lower emission profiles (e.g. chemical scrubbers or post-treatment systems) [28]. The third time period 2013-2015 is ongoing, and the current discussion concerns the extension of the system to the many waste water treatment plants that have biogas production.

3.2.6

Switzerland

One project performed in Switzerland looked for leakages on gas-conducting plant components at twelve plants by means of an IR-camera (Esders GasCam) and a portable methane analyzer (Esders Goliath) [29]. The leakages were not encapsulated and quantified. The data validation was limited to qualitative analysis. In Table 5 the detected weak points are dedicated to a relative significance. Table 5. Relative significance of the detected emission sources detected by the GasCam [29].

Emission source

Frequency

Reason

Emission potential

Connection of foil roof to tank wall

Very high

Constructional defect

Per leakage low  sum of leakages overall big

Pipe penetration

High


Big

36


Emission source

Frequency

Reason

Emission potential

Stirrer adjustment

High

Missing maintenance

Low

Open digester overflow

High


Medium

Bull eye

Medium


Low

Stirrer

Medium

Missing maintenance

Medium

Membrane

Medium

Material fatigue

Very big

Pressure relief vent

Low


Very big

Screw conveyor

Very low


Big

Inspection opening

Very low


Big

In the same project, methane emission measurements from open digestate storages were performed at three plants during four seasons. The measurements were performed with an open chamber system and complemented with batch tests in a laboratory (HBT – Hohenheimer biogas yield test) [29]. The measurements were not carried out directly in the storage. Instead the post digester, the open digestate storage and the solid digestate storage were sampled. The digestate samples were taken into the open chamber system (0.24 m3 volume) and each measurement lasted 5 days. The flow velocity inside the chamber was adjusted to the wind velocity (0.1 – 2 m s-1) over the open digestate storage. Simultaneously the samples were analyzed in the laboratory by the HBT. The results of the chamber measurements were extrapolated to calculate the methane loss. The results are presented in detail in Table 6, Table 7 and Table 8. Table 6. Methane emissions from digestate of the post digester [29].

Post digester

Methane emissions 3

Nm Sample 60 d

3

Methane production -1

Nm d tank

3

-1

Nm d

Methane emissions

Temperature in °C

% CH4-loss

Ambient

Digestate

Biogas plant 21 Jun2011

0.212

24.7

925

2.7

12.5

21.0

Sep2011

0.10

1.1

925

0.1

12.8

17.1

Dec2011

0.031

3.6

925

0.4

May2012

0.056

6.5

925

0.7

17.4

20.3

Biogas plant 13 Jul2011

0.028

2.3

578

0.4

20.1

25.2

Oct2011

0.011

0.9

578

0.2

13.2

18.6

Jan2012

0.018

1.4

578

0.3

1.7

11.6

Mar2012

0.024

2.0

578

0.3

12.7

16.5

Aug2011

0.058

3.2

784

0.4

20.4

21.6

Feb2012

0.000

0.0

784

0.0

0.0

0.9

Apr2012

0.026

1.5

784

0.2

12.1

12.5

Jun2012

0.075

4.2

784

0.5

15.8

19.3

Biogas plant 5

37


Post digester

Methane emissions

Average

0.046

Methane production

4.3

Methane emissions

Temperature in °C

0.5

Table 7. Methane emissions from digestate of the liquid digestate storage [29].

Liquid storage

Methane emissions 3

Methane production 3

-1

3

-1

Methane emissions

Temperature in °C

% CH4-loss

Ambien t

Digestate

Nm Sample 60 d

Nm d tank

Nm d

Jun2011

0.124

13.7

925

1.5

14.5

18.6

Sep2011

0.080

8.8

925

1.0

13.7

20.1

Dec2011

0.011

1.2

925

0.1

May2012

0.128

14.2

925

10.7

15.6

Biogas plant 21

1.5 Biogas plant 13

Jul2011

0.033

2.5

578

0.4

18.6

22.8

Oct2011

0.014

1.1

578

0.2

10.1

18.4

Jan2012

0.047

3.6

578

0.6

2.1

9.1

Mar2012

0.031

2.3

578

0.4

11.5

17.7

Biogas plant 5 Aug2011

0.183

9.7

784

1.2

18.9

19.6

Feb2012

0.000

0.0

784

0.0

7.0

7.7

Apr2012

0.038

2.0

784

0.3

9.0

12.7

Jun2012

0.026

1.4

784

0.2

15.8

18.7

Average

0.060

5.1

0.6

Table 8. Methane emissions from digestate of the solid digestate storage [29].

Solid storage

Methane emissions 3

Nm Sample 60 d

Methane production 3

-1

Nm d tank

3

-1

Nm d

Methane emissions

Temperature in °C

% CH4-loss

Ambien t

Digestate

Biogas plant 21 Jun2011

0.018

0.2

925

0.0

17.4

45.9

Sep2011

0.136

1.6

925

0.2

16.1

46.1

Dec2011

0.000

0.0

925

0.0

May2012

0.000

0.0

925

0.0

14.7

35.1

Jul2011

0.026

0.2

578

0.0

20.0

51.0

Nov2011

0.000

0.0

578

0.0

2.7

25.8

Jan2012

0.000

0.0

578

0.0

2.2

22.3

Mar2012

0.000

0.0

578

0.0

15.2

39.5.

23.4

55.0

Biogas plant 13

Biogas plant 5 Aug2011

0.200

1.1

784

0.1

38


Solid storage

Methane emissions

Methane production

Methane emissions

Temperature in °C

Feb2012

0.000

0.0

784

0.0

3.2

5.9

Apr2012

0.000

0.0

784

0.0

10.4

25.7

Jun2012

0.021

0.1

784

0.0

21.3

42.0

Average

0.033

0.3

0.0

The digestate from the post digester and the liquid digestate storage had comparable methane losses of about 0.5 – 0.6 % CH4 and the methane loss from the solid digestate was lower than 0.1 % CH4. The residual methane potential from the post digester varied between 3.8 and 14.5 %, from the liquid digestate between 3.3 and 13.5 % and from the solid digestate between 0.4 and 3 %. The measured methane emissions from the open chamber system were 2 – 24 % for the post digester, 3 – 37 % for the liquid digestate and 0 – 20 % for the solid digestate. In conclusion the results showed that the separation did not abate methane emissions from not gastight digestate storages. A specific project performed in Switzerland, but which is relevant to other countries as well, investigated the methane diffusion from single foil roofs [30]. The diffusion of the biogas storage foils was measured in a laboratory according to DIN 53380 (Testing of plastics - Determination of gas transmission rate - Part 2: Manometric method for testing of plastic films). The measurements were carried out with pure methane at a temperature of 41 °C, about 200 – 300 kPa absolute pressure and 0 % relative humidity. It must be pointed out that in reality biogas is water vapor saturated and consequently water condenses at the biogas storage foil. Furthermore biogas is a gas mixture of methane and carbon dioxide and the permeability coefficients of single gases in gas mixtures can possibly affect each other.

Methane diffusion rate in Ncm3/(m2∙d∙bar)

The single foil roofs from overall ten biogas plants and new foils from the manufacturers were investigated in a laboratory and their results are shown in Figure 9. 3000 2500 2000 1500 1000

500 0

Figure 9. Methane diffusion in Ncm3 m-2 d-1 bar-1 from single foil roofs [30].

39


The methane diffusion from new and unused EPDM-membranes (thickness 1.50 mm) is at between 1 690 and 2 190 Ncm³ m-2 d-1 bar-1 in contrast to the manufacturer information of 400 or 785 Ncm³ m-2 d-1 bar-1. A different temperature level between manufacturer information (23 °C) and the used temperature of 41 °C was also found. The methane diffusion from the investigated foil roofs was between 1 650 and 2 730 Ncm³ m-2 d-1 bar-1. Their layer thicknesses were between 0.8 and 1.5 mm. The methane diffusion showed a dependence of the layer thickness. With increasing thickness the methane diffusion was reduced. In case of filled biogas storages, the EPDM-foil expanded by up to 40 %. This could increase the methane diffusion calculation by approximately 24 %. Other influences, such as the used substrates in the biogas plant, or the age of the foils did not have verifiable influence on the methane diffusion.

3.2.7

Other countries

One project performed in Canada investigated a Canadian biogas plant over a whole year [31]. An average of 3.1 % CH4-loss during normal operation was measured using a remote sensing method (TDLAS and the software Windtrax). The results of the whole year are presented in Table 9. Table 9. Average seasonal fugitive emission rates as a percentage of seasonal methane production [31].

Autumn

Winter

Spring

Summer

Average

Methane production in m h

183

183

70

224

164

Methane emissions in % CH4

Normal operation

2.9

2.7

5.2

1.7

3.1

Flaring/ Venting

20

25

---

13

19

Maintenance

0.5

---

1.8

---

1.2

3

-1

40


4

Measurements at Tekniska verken in Linköping, Sweden

As a part of this project, comparative measurements of methane emissions were performed at the biogas plant in Linköping, Sweden, owned by Tekniska verken. Measurements were performed during the working week September 8th-12th in 2014 by six measurement teams, see Table 10. Table 10. Measurement teams and details on equipment and time for measurement.

Team

Approach

Analyzer

Meas. days

Report

AgroTech, DK

On-site

Photo-acoustic

Tuesday

Annex 1

DBFZ, DE (1)

On-site

GC-FID

Tue-Fri

Annex 2

DBFZ, DE (2)

Remote

TDLAS

Tue-Thu

Annex 3

DGC, DK

On-site

FID-Cutter

Tue-Wed

Annex 4

DTU, DK

Remote

CRDS

Tue-Wed

Annex 5

SP, SE

On-site

FID-Cutter

Mon-Tue

Annex 6

The complete measurement reports produced by the different teams are available as annexes to this report. These reports are generally detailed regarding the measurement methods and the application of methods at the Linköping plant. Hence, reporting of the results in this report does not go into much detail but rather tries to summarize and compare the results obtained by the different teams. For a basic understanding of the results presented later in this report it is however necessary to highlight some information concerning the measurement methods and analytical equipment applied by the different teams: 

AgroTech (on-site). AgroTech used an IR camera (FLIR GF 320) for the leak detection (together with DGC) and used a high volume sampler system during the emission measurements. Methane concentration measurements are made by a photo-acoustic analyser.



DBFZ (on-site). DBFZ used an IR camera (FLIR GF 320), a portable methane laser (GROWCON LaserMethane mini Gen2) and a portable biogas monitor (Geotech BM 2000) for the leak detection. For the emission measurements both direct measurements and open/closed chamber methods were used. Methane concentration measurements follow EN ISO 25139:2009 (GC-FID in laboratory).



DBFZ (remote). DBFZ used a TDLAS (GasFinder2.0, Boreal Laser Inc.) and a weather station with 3D sonic anemometer (Model 81000, RM Young) to measure path averaged concentrations and meteorological conditions in the

41


downwind plume of the site. Inverse dispersion modelling was performed using the software Windtrax [32]. 

DGC (on-site). DGC used an IR camera (FLIR GF 320) for the leak detection (together with AgroTech) and used a high volume sampler system during the emission measurements. Methane concentration measurements follow EN ISO 25140:2010 (FID with Cutter).



DTU (remote). DTU measured downwind plume concentrations of methane and the tracer gas acetylene using a mobile analytical platform holding a fast and high sensitive cavity ring down spectrometer (CRDS) capable of detecting methane and acetylene concentrations down to ppb level every second, a GPS and a weather station.



SP (on-site). SP used a portable leak detection instrument (Sewerin EX-TEC PM4) for the leak detection and followed the guidelines in the Swedish Voluntary Agreement system (see section 3.2.5) during the emission measurements. In short, only well-defined and systematic emission sources are covered by the quantification method, excluding leaks and other diffuse sources. Direct measurements of the gas flow and methane content are made. Methane concentration measurements follow EN ISO 25140:2010 (FID with Cutter).

4.1

PLANT DESCRIPTION

The Åby biogas plant is one of Sweden’s largest biogas plants, owned and operated by Tekniska Verken i Linköping. It is situated in Linköping 200 km south-west of Stockholm and has been in operation since 1996. The plant has an annual permit of treating 100,000 tons of organic waste, and the annual production of raw biogas was about 17,000,000 Nm3 in 2013. Besides biogas the plant produces 80,000 tons of wet bio fertilizer, which is sold to farmers. Table 11 shows the distribution of the substrates treated at the plant. Table 11. Substrate input in 2013.

Type

Proportion (% weight)

Slaughterhouse cat. 3

20

Food Industry

35

Alcohol

0.5

Thin stillage

1.5

Fat

2

Glycerol

1

Food waste

40

The plant production layout comprises receiving and pre-treatment facilities, four digesters, gas storage/upgrading and biofilter. The layout is shown in Figure 10.

42


Figure 10. Layout of Åby biogas plant, Tekniska Verken i Linköping.

4.2

LEAK DETECTION

Leak detection was performed on Monday afternoon by all the on-site measurement teams. AgroTech and DGC performed it together as one team. The three different teams performed leak detection separately without influencing each other. Some already known emission points such as digestate storage, biofilter release and gas upgrading CO2 release were not part of the leak detection campaign, the focus was on trying to identify other sources of emission such as equipment leaks or emissions from storage tanks. Results were compared at a meeting during Monday evening and then the teams decided which emission points should be measured during the following days. Table 12. Results from leak detection campaign performed on Monday.

Leakage

DBFZ

AgroTech/DGC

SP

Upgrading unit, meas. device inside the right compressor unit

Detected

Detected

Not detected, does not look for leaks inside ventilated buildings

Upgrading unit, safety valve pipe on roof

Detected

Detected

Not detected

Carbon filter house

Detected

Detected

Detected

Hygienization tank, manhole and pipe

Not detected

Detected

Detected

Hygienization tank, one-hour holding tank

Not detected

Detected

Not detected

Homogenization tank

Detected

Detected

Not detected

43


Leakage

DBFZ

AgroTech/DGC

SP

Digester no 3

Not detected

Not detected

A few minor gas leaks were detected on the roof

A brief outline of the results from the leak detection campaign is presented in Table 12. Pictures of the identified leaks are shown in Figure 11 and their placement on the plant is shown in Figure 12. Leakage # 1

Leakage # 4

Upgrading unit – ventilation duct 1

Carbon filter house

(facing north – away from the biogas plant)

ventilation duct leading from the container

Leakage # 2

Leakage # 5

Upgrading unit – ventilation duct 2

Carbon filter house – gas booster F22

(facing east towards digester 4)

Leakage # 3

Leakage # 6

Upgrading unit – safety valve pipe leading above

Carbon filter house – gas booster F21

the container

44


Figure 11 a). Overview of detected leakages, arrows indicate precise location of the leakages (taken from DGC report).

Leakage # 7

Leakage # 10

Carbon filter house – gas filter

Homogenisation tank swan neck

Leakage # 8

Leakage # 11

Hygienisation tank

Homogenisation tank

man hole at the top of the tank

hole in the concrete roof

Leakage # 9

Leakage # 12

Hygienisation tank

Hygienisation tank (one-hour holding tank)

overloading valve

swan neck at the top of the tank

45


Figure 11 b). Overview of detected leakages, arrows indicate precise location of the leakages (taken from DGC report).

46


Figure 12. Overview of the plant and the detected leakages + other emissions sources (taken from DBFZ on-site report).

47


Results from the leak detection campaign depend on the detection equipment used, the experience of the personnel performing the leak detection and variations in plant conditions. SP did not find the leaking safety valve pipe on the roof of the upgrading unit, since this area was not investigated by them. SP uses a portable leak detection device which needs to be situated in the emission point to allow for detection, whereas the other two teams use IR cameras that are able to detect leaks from a distance. Leak detection is not performed by the measurement teams in the Swedish Voluntary Agreement system (see section 3.2.5) but the plant owner needs to have routines and equipment to perform it themselves. On the other hand SP detected some minor leaks on top of digester no. 3 that the other teams did not find due to the relatively large detection limit of the IR cameras. DBFZ did not detect the leaks at the hygienization tank and SP did not detect the leaks at the homogenization tank. Neither of these contributes to any great deal to the total emission measured later on (see section 4.3). AgroTech/DGC detected a leakage with the IR camera at a swan neck exit on top of a one-hour holding tank which is part of the hygienization process. They were however in doubt whether the leakage detected with the IR camera was methane or just emitted heat from the hot tank. It can be hard to differ between methane and heat emission with the IR camera. The emission was investigated later during the week and was not found to be emitting methane. The conclusion of the results from the leak detection campaign is that an IR camera is a very good tool to be able to find leaks at a biogas plant in an effective way, but since it is hard to differ between methane and heat with the cameras, a portable leak detection instrument is a necessary complement. It must be pointed out that a portable leak detector should preferably use a methane specific measuring principle. Instrumentation which is based on e.g. semiconductors often has cross sensitivity to hydrogen sulfide. Certain sources, such as the hygienization or the homogenization tank are then possibly overestimated concerning methane concentration measurements. For example, at the flange of the hygienization tank SP measured about 10 vol-% (thermal conductivity sensor) and DBFZ measured only 0.1 vol-% CH4 (IR analyzer).

48


4.3

ON-SITE EMISSION MEASUREMENTS

With all the reports from the teams finally available in January 2015, a comparison spread-sheet was made and circulated to the teams. The spreadsheet was discussed thoroughly during a telephone meeting with all the teams, and the revised version of the spread-sheet following this meeting is the basis of the results and how they are presented in this chapter. Individual results here are given in the unit g/h, whereas the emission rates are always given in the unit kg/h in other parts of this report. Table 13 indicates at what times the respective teams performed measurements at the individual sources. Table 13. Measurement days and times for the respective emission sources.

Emission source

AgroTech

DBFZ

DGC

SP

Homogenization tank

Tues. 16:1017:00

Tues. 14:45-16:00

Wed. 10:30-12:05

N/A

Hygienization tank

Tues. 15:3516:05

Thur. 08:00-13:00

Wed. 15:30-18:15

N/A

Biofilter

N/A

Tues. 11:45-13:00

N/A

Mon. 14:00-16:00

Digestate storage

N/A

Tues. 16:00-18:45

N/A

Tues. 13:00-16:45

Activated carbon filter building

Tues. 17:4018:40

Wed. 08:00-13:00

Tues. 16:00-19:15

Tues. 15:00

Chemical scrubber, CO2 release

N/A

Wed. 15:00-16:00

N/A

Tues. 8:55-10:20

Chemical scrubber, compr. buildings

Tues. 10:5512:50

Wed. 17:00-19:00

Tues. 11:30-12:50

Tues. 16:30

Chemical scrubber, prs. relief vents

Tues. 11:2012:15

Thur. 14:00-15:00

Tues. 14:20-15:00

N/A

N/A

N/A

N/A

Tues.

Analysis instruments on site

4.3.1

Homogenization and hygienization tanks

SP did not measure these sources. The estimated methane mass flows were considered too low based on methane concentrations given by their portable leak detection instrument and not detectable gas flow. The preliminary measurements did not motivate performing proper measurements, also given that the nature of these sources would give values with a very high uncertainty and/or involve safety risks. Furthermore SP does not have the equipment or methods available that are necessary to perform measurements from the leaking flange (man-hole) on the hygienization tank. AgroTech argues that they would normally not include the homogenization tank in a measurement campaign since it is not part of the digestion process and also since their measurement method is not suitable for releases with low or even non-existing pressure difference between the emission source and the ambient air (see discussion

49


below). They did however perform the measurements for this project to compare their results with the other teams. The results from the measurements are presented in Table 14. Table 14. Emission measurements in homogenization and hygienization tanks

Emission source (g CH4/h)

AgroTech

DBFZ

DGC

SP

Homogenization tank (sum)

104

0.46

6

N/A

- hole

86

0.45

3

N/A

- pipe

18

0.01

3

N/A

Hygienization tank (sum)

88

16

96

N/A

- flange

7

10

6

N/A

- pipe

81

6

90

N/A

Total percentage methane loss

0.02 %

0.002 %

0.01 %

N/A

The results for the homogenization tank differ with a factor of 100 between the lowest and highest values. The homogenization tank is characterized by open access (holes and vent pipe) from the surrounding atmosphere to the confined tank volume meaning the tank “breathes” with the impact of atmospheric pressure, wind, stirring, supply of waste etc.. Therefore the pressure difference between the surroundings and the tank is close to zero. The larger the pressure difference between a volume of confined biogas and the atmosphere, the more precise the measurement result by the high volume sampler method (AgroTech and DGC) will be. Using the high volume sampler method will, at a source like this, always be a trade-off between affecting the source as little as possible with the risk of not detecting the entire emission or using a too high sample flow causing an overestimation of the emission by sucking out gas from the tank. At the time of DGC measuring, 10 September 10:30-12:05, there were relatively strong winds and there was some heavy turbulence at the sample location. In order to minimize the impact on the measurement DGC built a “tent” to cover the hole and pipe. The air/sample flow was chosen as low as possible to continuously securing a positive flow across the sample plane of the tent (no back flow with loss of methane emission caused by wind turbulence). The velocity was checked by a vane anemometer covering the entire cross sectional area (grid measurements). The result was a sample flow of approx. 500 m3/h simulating a velocity of 0,4-0,5 m/s across the leakages. Figure 13 shows how the tank breathes. DGC measured emissions ranging from zero (back ground level) to approx. 70 ppm. The fact that emissions occasionally fall to zero during the entire sample period indicates, that the sample flow applied is not affecting the emission excessively. As for the swan neck (pipe) the emissions never fall to background level but vary between 3 and 30 ppm, see Figure 14. One possible reason is that the sample flow was too large creating a vacuum resulting in sucking out gas from the tank. But this could also be caused by another explanation. The pipe can act as a chimney which creates a natural draft conveying a small flow of gas from the tank. The emission rate from a source where practically no pressure difference exists between atmosphere and the confined biogas volume is very prone to impacts by the high volume sampler method. It is believed that AgroTech overestimated the emission due to the difficulties described and DGC thinks that there is a small probability that

50


even their result could be overestimated and the measured emission may have been lower if a lower sample flow had been applied. It is further believed that DBFZ underestimated the emission since they used a closed chamber for their measurement (Figure 15), and this would also have had a great impact on the way that this emission source behaved.

Figure 13. DGC results and sampling set-up at homogenization tank – hole.

Figure 14. DGC results and sampling set-up at homogenization tank – pipe.

Figure 15. DBFZ sampling set-up at homogenization tank – hole.

For the hygienization tank the reported values are more in comparison. Strong variations in methane concentration were observed by the teams during their respective measurements. To finalize the discussion on these emission sources, we would like to strongly point to the fact that they contribute very little to the total emission of the plant, and one team (SP) did not even perform measurements due to this fact.

51


4.3.2

Biofilter

AgroTech and DGC did not measure in this source since they do not have the equipment available that is necessary. The results from the measurements are presented in Table 15. Table 15. Emission measurements in biofilter release.

Biofilter

AgroTech

DBFZ

DGC

SP

N/A

152

N/A

204

N/A

126

N/A

57

(gas flow Nm /h)

N/A

1 209

N/A

3 564


N/A

0.02 %

N/A

0.02 %

Emission rate (g CH4/h) 3

(methane concentration mg/Nm ) 3

DBFZ placed a plastic foil on the surface of the biofilter and took gas samples from it to analyze the methane concentration (Figure 16). Gas flow was measured in the pipe leading to the biofilter. SP measured both methane concentration and gas flow in the pipe leading to the biofilter. This is done according to the Swedish Voluntary Agreement system since it is there defined that there are no available results that show that biofilters would have any effect on the methane concentration. Biofilters will (at best) have an effect on odorous substances.

Figure 16. DBFZ sampling set-up at biofilter.

Despite the different approaches the mass flows reported by DBFZ and SP are in good comparison, but there is however a large difference in the underlying measurements of methane concentration and gas flow (Table 15). DBFZ performed their measurement on Tuesday when the measured gas flow was ca 4 m/s, but when they quickly checked the gas flow again on Friday it was then about 10 m/s. It is therefore believed that the gas flow to the biofilter varies quite a lot over time, but this variation does not seem to have much impact on the emitted mass flow of methane.

52


4.3.3

Digestate storage

AgroTech and DGC did not measure in this source since they do not have the equipment available that is necessary. The results from the measurements are presented in Table 16. Table 16. Emission measurements in digestate storage.

Digestate storage

AgroTech

DBFZ

DGC

SP

Emission rate (g CH4/h)

N/A

4 382

N/A

7 335


N/A

0.51 %

N/A

0.85 %

Looking at the reported results there seems to be a large difference in measurement results, but looking at the individual measurements it is revealed that this difference is not due to a difference in the measurement results but rather due to assumptions and the way the results are treated. DBFZ performed measurements with a closed chamber at four points on the surface of the digestate storage. They chose one of the points where the surface layer was considered to be rigid, whereas the other three chosen points had a cracked surface layer. Further the assumption was made that 50 % of the total surface has a cracked surface layer and 50 % has a rigid surface layer, and the measurement results were corrected relative to this assumption. SP on the other hand did not purposely choose their four sampling points due to the appearance of the surface layer, but rather chose the points on random. SP reported the mean value of the individual results from the four sampling points. SP used an open chamber for their measurement. Looking at the individual measurement results in Figure 17, the DBFZ point on the rigid surface layer appears as an outlier with a notably lower value than the other measurements. All the other seven measurements are in agreement. Figure 18 shows the digestate storage tank and the appearance of its surface layer. The open chamber used by SP can also be seen in this figure (if the reader looks closely), which also gives an indication of the task necessary to perform sampling in many sampling points. Four sampling points – all near the wall of the tank for practical reasons – could probably not be regarded as representative of the whole surface layer.

53


10 Methane emission (g/m2 h)

9 8 7 6

SP

5

DBFZ

4 3 2 1 0

Figure 17. DBFZ and SP individual measurement results at digestate storage (4 sampling points each).

Figure 18. Digestate storage tank and the appearance of the surface layer.

54


4.3.4

Activated carbon filter building

The results from the measurements are presented in Table 17. Table 17. Emission measurements in activated carbon filter building.


AgroTech

DBFZ

DGC

SP

Activated carbon filter

97

138

133

80

- leakage 1st compressor (F22)

75

68

111

N/A

- leakage 2nd compressor (F21)

17

63

56

N/A

- leakage flange

5

4

9

N/A

- calculated sum of individual leaks

97

135

176

N/A


0.01 %

0.02 %

0.02 %

0.01 %

In the activated carbon filter building, three different leakages were detected. AgroTech applied their method on the individual leaks whereas SP measured the total emission from the building in the ventilation release. DBFZ and DGC used both approaches with very good agreement. DGC overestimated the individual leaks due to contaminated background air in the building. DBFZ used fresh air from outside the building for their individual measurements, giving an almost perfect agreement between individual leakages measurement and total emission measurement in the ventilation. It is believed that SP underestimated the gas flow in the ventilation since this was measured with a hotwire anemometer directly in the ventilation opening. DGC used the high volume sampler method also in the ventilation opening.

4.3.5

Chemical scrubber, CO2 release

AgroTech and DGC did not measure in this source. The results from the measurements are presented in Table 18. Table 18. Emission measurements in chemical scrubber, CO2 release.

CO2 release

AgroTech

DBFZ

DGC

SP

N/A

N/A

N/A

85

(methane concentration mg/Nm )

N/A

273

N/A

121


N/A

N/A

N/A

0.01 %

Emission rate (g CH4/h) 3

DBFZ performed measurements on Wednesday when the chemical scrubber was running on ca 50 % load and do not have the data available from the plant to be able to calculate the mass flow of methane. SP performed measurements on Tuesday when the chemical scrubber was running on full load. The measurement point available for sampling is not suitable for flow measurement. To be able to calculate the emissions rate SP instead used the values of methane content and produced gas (Nm3/h) from the plant, which is standard practice. But since cooling air from the compressors is also released at this measurement point at the Linköping plant the calculation will not be correct in this case.

55


4.3.6

Chemical scrubber, compressor buildings

The results from the measurements are presented in Table 19. Table 19. Emission measurements in chemical scrubber, compressor buildings.


AgroTech

DBFZ

DGC

SP

84

198

285

363

21

44

59

81

5 100

4 476

4 944

4 475

105

66

159

98

13

12

25

22

(gas flow Nm /h)

6 330

5 751

6 740

4 535


0.02 %

0.03 %

0.05 %

0.05 %

Right compressor building 3


(gas flow Nm /h) Left compressor building 3


These sources illustrate the influence of both the methane concentration measurement and the gas flow measurement to the end result. All teams but SP use background values of methane concentration to compensate the source measurements. In this particular case this would be the methane concentration of the air that is led in to the respective compressor buildings, and this concentration is then raised by any leakages in equipment inside the building, before the air is emitted. For these particular sources DBFZ and DGC used background values of 2 ppm, whereas AgroTech used values of 34 ppm and 12 ppm respectively. It is believed that AgroTech’s measurement of background values was contaminated by the high emission of the leaking safety valve on the roof of the building. Particularly for the left compressor building, this correction for the background value has a relatively large influence on the reported result. Taking the above in consideration when looking at the methane concentration measurement results in Table 19, it shows a good relation between teams for the left compressor building but with some larger variations for the right compressor building. It is believed that there was some variation in the concentration over time, and it might very well also be due to the leaking safety valve nearby. SP use fan data for the gas flow values and since the fans are identical in the two buildings they report the same value for both buildings (slight difference due to temperatures). The fan data is in good agreement with the flow measurements performed by the other three teams for the right compressor building, but all measurements are consistently higher for the left compressor building for which we have no good explanation.

56


4.3.7

Chemical scrubber, pressure relief vents

SP was not able to perform measurements on the leaking pressure relief vent, since the compressor was not running on the Wednesday when measurements were going to be performed. DBFZ performed measurements on the Thursday and at this time the plant personnel had fixed the leaking safety valve. Instead DBFZ detected another leaking relief vent (not covered by the emission source list) and performed measurements on it. The results from the measurements are presented in Table 20. Table 20. Emission measurements in chemical scrubber, pressure relief vents.


AgroTech

DBFZ

DGC

SP

Prs. relief vent, left building

6 010

N/A

9 159

N/A

Prs. relief vent, between buildings

N/A

317

N/A

N/A


0.70 %

0.04 %

1.1 %

N/A

The only comparable results are between AgroTech and DGC for the pressure relief vent on the left building. The difference in reported numbers is quite large. DGC are quite confident regarding the result obtained because two measurements were performed at different sample flows (duplicate determination): 

Measurement #1: Sampling hose was mounted directly on the vent pipe and sufficient sample flow controlled by checking for back flow of leaking biogas.



Measurement #2: Sampling was mounted on the hose and the hood was equipped with plastic skirts to prevent wind turbulence causing leaking biogas to escape. This method resulted in less pressure loss in the sampling system and consequently a larger sample flow and lower CH4 concentration.

Measurement #2 is the one reported by DGC. Table 21 indicates that the CH4 emission measured by the two different sample flows is almost identical. Table 21. DGC results for the pressure relief vent on the left building.

Measurement

Sample flow (Nm3/h)

CH4 (ppm)

Methane flow (g/h)

#1 (13:15 – 14:08)

507

24 650

9 000

#2 (14:22 – 15:03)

745

17 067

9 159

4.3.8

Analysis instruments on site

SP made notes of the sample flows passing through the analysis instruments that are permanently installed at the site, since this is normal practice in the Swedish Voluntary Agreement System. No other team did this. The sample flow through the one analyzer installed in the activated carbon filter building is however included in the DBFZ and DGC measurements from this source. The results from the measurements are presented in Table 22. Table 22. Emission measurements in analysis instruments.

Analysis instruments

AgroTech

DBFZ

DGC

SP

Emission rate (g CH4/h)

N/A

N/A

N/A

41


N/A

N/A

N/A

measurements of methane emissions from biogas production

measurements of methane emissions from biogas production

Suggest Documents

Production of biogas (methane and hydrogen) from

BIOGAS PRODUCTION BY ENCAPSULATED METHANE ...

Measurements of methane emissions at natural gas production sites in ...

Methane production and bubble emissions from

Reducing Methane Emissions from Production Wells: Reduced ...

Measurements of methane emissions at natural gas production sites in ...

Potential of biogas and methane production from anaerobic digestion ...

Direct measurements of methane emissions from abandoned oil and ...

Short communication: Measurements of methane emissions from feed

Measurements of methane emissions from ... - Atmos. Meas. Tech

Direct measurements of methane emissions from abandoned oil and ...

Production of Methane Emissions from Ruminant Husbandry: A Review

Quantifying Methane Emissions from Reservoirs

Methane production and bubble emissions from ... - Semantic Scholar

Methane production and bubble emissions from ... - Semantic Scholar

Chamber Size Effects on Methane Emissions from Rice Production

10.6 MEASUREMENT OF METHANE EMISSIONS FROM CATTLE ...

Biogas Production from Biological Sludge

Potential Biogas Production from Artichoke

biogas production from wastewater sludge

Fugitive emissions of methane from abandoned ... - ReFINE

Effect of Feed Loading on Biogas Methane Production in Batch ...

OPTIMISATION OF BIOGAS PRODUCTION FROM ... - CiteSeerX

Technical Possibilities of Biogas Production from